Shubham Singh

About

Backend engineer at Cerebrone.ai building cloud-native Flask microservices for an AI-driven automation platform on GCP (Cloud Run/App Engine), including dedicated inference services integrating OpenAI and internal ML pipelines. Demonstrated strong performance and scalability wins across Postgres/SQLAlchemy optimization, multi-tenant (healthcare/HIPAA-grade) data isolation, and high-throughput background processing with Celery/Redis/RabbitMQ, with multiple quantified latency/CPU/throughput improvements.

Experience

Education

Key Strengths

• Designed and deployed Flask microservices on GCP (Cloud Run/App Engine) with Docker for autoscaling
• Built stateless services using JWT to enable horizontal scaling
• Created dedicated AI inference microservice with async communication to ML pipelines enabling zero-downtime model updates
• PostgreSQL/SQLAlchemy performance tuning (indexes, eager loading, keyset pagination, JSONB, partial indexes, EXPLAIN ANALYZE)
• Reduced analytics query latency from ~1.8s to <300ms and cut Postgres CPU ~40%
• Implemented hybrid multi-tenant isolation (schema-per-tenant + RLS + partitioning + read replicas) improving API latency ~35% and easing compliance audits
• Optimized background processing with Celery + Redis/RabbitMQ, batching, caching, and worker scaling; cut API latency ~2s to ~400ms and increased throughput 4–5x
• Production reliability practices: monitoring/logging with Stackdriver/Cloud Logging and automated tests with PyTest/Postman
• Accelerated model/feature delivery (reported ~25% reduction in model integration/feature delivery time) via automation and architecture